FITAI - Artificial Intelligence for GNSS post-fit residual and position joint analysis. A plugin for Jason, Rokubun cloud GNSS processing service
- Title: FITAI - Artificial Intelligence for GNSS post-fit residual and position joint analysis. A plugin for Jason, Rokubun cloud GNSS processing service
- Funder: ESA
- Call: Open Discovery Ideas Channel
- Budget: 175.000 €
- Partners: Rokubun
- Duration: 18 months
Machine Learning algorithms in the field of GNSS are usually trained using physical models of the different aspects that make up the GNSS signal (the geometric range, ionospheric and tropospheric delay, orbits, clocks, etc.) and positioning filters such as Kalman filter that estimate a state (position, velocity and time) based on a set of observations. However, after the estimation has been obtained, little attention is paid to the ‘post-fit residuals’ – the values that indicate how well the input data ‘agrees’ with the estimated state.
The FITAI project will use these post-fit residuals and positioning errors to develop a processing stage based on an ESA patent for Rokubun's Jason cloud GNSS service. This will improve the accuracy of navigational satellite position estimates, particularly in challenging scenarios such as urban canyons.
